The work presented here encompasses two lines of research broadly concerned with the way people think of themselves and of others in the context of knowledge and expertise. The first line of research focuses on self-judgments of knowledge. People are often required to evaluate their own knowledge in the service of various decisions; how is this evaluation made, and when does it go awry? One way to study overestimation of knowledge is by asking people to assess their knowledge of terms, places, people, etc. that were invented for the sake of research and do not exist. This approach helps reveal when and why people mistakenly believe themselves to have knowledge that they could not possibly have, a phenomenon called overclaiming. Across two papers, I demonstrate that people are more likely to overclaim knowledge about nonexistent terms when they perceive themselves to be overall knowledgeable about the topic from which the terms are drawn. However, self-perceptions of overall knowledge can be dissociated from true knowledge. I also find that the more people are genuinely knowledgeable about a topic, they less likely they are to overclaim – i.e., to mistakenly believe they have knowledge of nonexistent terms related to their topic of expertise.